Walk-based diversification for data summarization

Abstract

Due to the large amount of data stored in current information systems, new strategies are required in order to extract useful information from databases. Hereupon, data summarization is an interesting process that allows reducing a large database maintaining just the relevant parts of the whole collection. In this study, we propose a new approach for data summarization based on a recently proposed tourist walk diversification method. This approach allows setting two ways of selecting elements considering density and hyper volume of each class. In order to evaluate the proposed approach, we compared it with two known methods of the literature considering one real world dataset and one artificial dataset. The artificial dataset was created considering different data distribution aspects. The conducted experiments outcomes demonstrate that our proposed data summarization approach is a promising alternative for addressing the problem of selecting elements from large databases considering different aspects of distribution.